hushem_1x_deit_tiny_rms_lr0001_fold5

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 2.1537
  • Accuracy: 0.5610

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 0.0001
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 6 1.5741 0.2683
1.8922 2.0 12 1.3978 0.2683
1.8922 3.0 18 1.4032 0.2439
1.5101 4.0 24 1.4021 0.2683
1.38 5.0 30 2.1528 0.2439
1.38 6.0 36 1.4141 0.2439
1.4096 7.0 42 1.2484 0.4390
1.4096 8.0 48 1.2607 0.4390
1.2381 9.0 54 0.9950 0.5366
1.1539 10.0 60 1.0350 0.5610
1.1539 11.0 66 1.2716 0.3415
0.9039 12.0 72 1.0596 0.5854
0.9039 13.0 78 1.5972 0.4146
0.6191 14.0 84 1.9855 0.4390
0.4358 15.0 90 1.2403 0.4878
0.4358 16.0 96 2.3374 0.4390
0.2291 17.0 102 1.5475 0.4390
0.2291 18.0 108 1.2789 0.6341
0.1203 19.0 114 1.8441 0.4390
0.0604 20.0 120 1.7948 0.4878
0.0604 21.0 126 2.0211 0.4634
0.0322 22.0 132 1.8178 0.5366
0.0322 23.0 138 2.0950 0.4878
0.017 24.0 144 2.0410 0.5122
0.0011 25.0 150 2.0405 0.5122
0.0011 26.0 156 2.0495 0.5122
0.0007 27.0 162 2.0594 0.5122
0.0007 28.0 168 2.0747 0.5122
0.0006 29.0 174 2.0825 0.5610
0.0005 30.0 180 2.0915 0.5610
0.0005 31.0 186 2.1017 0.5610
0.0004 32.0 192 2.1110 0.5610
0.0004 33.0 198 2.1199 0.5610
0.0004 34.0 204 2.1276 0.5610
0.0004 35.0 210 2.1335 0.5610
0.0004 36.0 216 2.1398 0.5610
0.0004 37.0 222 2.1439 0.5610
0.0004 38.0 228 2.1473 0.5610
0.0003 39.0 234 2.1497 0.5610
0.0003 40.0 240 2.1519 0.5610
0.0003 41.0 246 2.1532 0.5610
0.0003 42.0 252 2.1537 0.5610
0.0003 43.0 258 2.1537 0.5610
0.0003 44.0 264 2.1537 0.5610
0.0003 45.0 270 2.1537 0.5610
0.0003 46.0 276 2.1537 0.5610
0.0003 47.0 282 2.1537 0.5610
0.0003 48.0 288 2.1537 0.5610
0.0003 49.0 294 2.1537 0.5610
0.0003 50.0 300 2.1537 0.5610

Framework versions

  • Transformers 4.35.0
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1
Downloads last month
22
Safetensors
Model size
5.53M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for hkivancoral/hushem_1x_deit_tiny_rms_lr0001_fold5

Finetuned
(363)
this model

Evaluation results